June 



19^ 



NATURE 



355 



" trilobites," the adult condition of which is still un- 

 known. Mr. Gahan suggests that the larval stage 

 may be permanently retained in the females. 

 \ Extinct North American horses, all referable to the 

 modern genus, form the subject of a paper by Mr. 

 O. P. Hay, published as No. 1969 (vol. xliv., pp. 

 569-594) of the Proceedings of the U.S. National 

 Museum. Four species are described as new, two 

 of these being based on teeth alone, while each of the 

 other two is represented by the skull. No. 1975 

 (vol. xliv., pp. 649-654) of the same serial is devoted 

 to the description of a remarkably fine skeleton of a 

 zeuglodont lately set up in the museum. For these 

 primitive whales Mr. J. W. Gidley, the author of the 

 article, revives the extremely inappropriate name 

 Basilosaurus, despite the fact that it was replaced by 

 its sponsor, Sir R. Owen, by Zeuglodon when the 

 mammalian nature of the remains became evident. 



We have received reprints of two papers by Mr. 

 W. E. Collinge on the relation of wild birds to 

 forestry and on the destruction and dispersal of weed 



eds by wild birds. In the former it is pointed out 

 that but few wild birds are directly hurtful to forests 

 in this country, the majority of species found in or 

 near forests being distinctly beneficial owing to 

 their destruction of insects and small mammals. 

 Attention is directed to the importance of providing 

 nesting-boxes in forests for the insectivorous birds. 

 Merely to protect these birds is not sufficient ; their 

 multiplication must also be looked after. In the 

 second paper many interesting details are given as to 

 the frequencv with which birds completely destroy 

 even hard fruits and seeds, but there are so many 

 cases in which the seeds pass through the bird's 

 alimentary canal without being injured that the 

 author cannot regard the seed-eating birds as a class 

 as being beneficial — on the whole, they appear to act 

 as distributors of weed seeds to a much larger extent 

 than is generally supposed. 



At the commencement of a very interesting article 

 on our present knowledge of the earliest quadrupedal 

 vertebrates (Tetrapoda), Prof. F. Broili points out 

 that the occurrence of remains of land plants in the 

 Lower Silurian of North America and the Upper 

 Silurian of Kellerwald and the Hartz, coupled with 

 their occasional presence in the Lower Devonian, 

 may be taken as strong presumptive evidence of the 

 existence of a contemporary vertebrate land-fauna. 

 The first actual evidence of such an earlv fauna does 

 not occur, however, until the Upper Devonian, and 

 then only in the shape of the remarkable footprint 

 from Warren County, Pennsylvania, to which O. C. 

 Marsh gave the name Thinopus.- Very noteworthy 

 is the fact that this print represents a relatively large 

 animal, vastly superior in point of size to the tinv 

 salamanders of the succeeding Lower Carboniferous 

 epoch, although equalled in this respect by Eosaurus 

 of the Upper Carboniferous of Nova Scotia. Its 

 evidence appears, however, quite indisputable, and 

 we must therefore assume the Tetrapoda to have 

 attained a relatively high degree of development in 

 the Devonian, and thus to have been well repre- 

 sented in the Silurian. This being so, the Carbon- 

 NO. 2275. VOL. 91] 



iferous and Permo-Triassic forms, interesting and in 

 many respects generalised as they are, cannot be 

 regarded as the earliest types, and, consequently, any 

 attempts to formulate the phylogeny of the group 

 must be to a great extent premature. The author 

 has much to say regarding the structure and affinities 

 of labyrinthodonts and anomodonts, but space permits 

 only of the remark that he regards most of these, not 

 even excluding dicynodonts, as amphibious. 



We have received three numbers of the Bolletiino 

 of the Italian Seismological Society, forming a volume 

 of more than 300 pages, and containing notices of the 

 earthquakes recorded in Italy during the first half 

 of the year 1909. The total number of entries is 

 726, all but seventy of which refer to shocks that 

 originated within the area of Italy. This number, 

 which is much greater than usual, is partly due to the 

 frequency of the after-shocks of the Messina earth- 

 quake of December 28, 1908. Dr. Martinelli, the 

 editor of the volume, connects 152 shocks with this 

 great earthquake, but many shocks recorded at Mes- 

 sina and other isolated places, if more fully known, 

 might have to be placed in the same category. 



The Commonwealth Meteorologist has recently 

 issued sheet maps showing the mean temperature 

 and mean rainfall of Australia for the separate months 

 and for the year. These maps confirm in the main 

 those published in the " Atlas of Meteorology," so 

 far as regards the south-eastern portion of the larger 

 island ; they extend the information over the re- 

 mainder of the continent, and thus include Tasmania. 

 The earlier maps, based on the Challenger results, 

 show temperatures for six alternate months only. 

 The most striking points of difference occur in the 

 annual maps. The annual isotherms in the atlas map 

 sweep southwards between the coasts; the new maps 

 show roughly a sweep to the north. The annual 

 isohyets on the new maps show much greater detail 

 for the north and east coasts, and indicate that 

 parts of the north coast receive more than 60 in. and 

 Geraldton district on the Queensland coast more than 

 140 in. annually. The new monthly rainfall maps 

 indicate clearly the winter rains in the south and the 

 summer rains in the north. 



According to the resume of the communications 

 made to the French Physical Society at the meeting 

 on April 18, a large proportion of the evening was 

 devoted to a discussion of the results of recent deter- 

 minations of the radiation constants, <r of Stefan's 

 law and c of Wien's and Planck's laws. One of 

 the most recent and most accurate determinations of 

 c is that of Warburg, Leithauser, Hupka, and 

 Muller, made at the Reichsanstalt, and reported in 

 the April number of the Annalen der Physik. Accord- 

 ing to this determination c has the value 1437 

 centimetre-degrees, a number which differs little from 

 the mean of the more modern of the previous deter- 

 minations. The determinations of tr differ much more 

 seriously from each other. The oldest values lie 

 between 50 and 5-5, while almost all the newer 

 values are above 55, those of Valentiner and Westphal 

 being 5-58 and 5-54 respectively, while M. Fe>y has 

 in three sets of experiments obtained values above 



